Newborn infants prefer the sound of their own mother's versus another mother's voice. Little is known about the development of this selective response and the cues which are salient to the infant in identification of mother's voice. Intrauterine experience is a likely precursor, and the response may reflect a preference for a familiar sound or perceptual bias resulting from early exposure. Recordings made in the amniotic environment indicate emergence mothers's voice over other voices against a background of vascular and digestive noise. In a recent study using a nonnutritive sucking choice procedure, two-day-olds preferred the sound of mother's simulated in utero voice versus a normal recording of her voice. There was evidence to suggest that this preference may reverse in the first postnatal week with extrauterine experience (Moon & Fifer, 1986). The main goal of the proposed research is to investigate early perception of the speech signal, and in particular, to characterize responding to three cues which are likely to be salient to newborns. The three cues are prosody filtering, and the unique qualities of mother's vocal tract. These cue will be systematically combined in pairs of stimuli which are expected to elicit a preference by the infant. A change in a salient cue is expected to result in a change in preference. Conclusions about the relative salience of the three cues will be based upon results from individual experiments, and inferences will be made from patterns of results. A further goal is to investigate the generality of responding to cues in the intrauterine speech signal. Non-speech acoustic stimuli will be presented with intra- versus etrauterine features. The results of the proposed project will have implications for early speech perception and for early cognitive capacities.